Key Moments
Michael Seibel - How to Plan an MVP
Y CombinatorKey Moments
Launch a bad product quickly to get your first users and feedback, as even billion-dollar companies like Airbnb and Stripe started with flawed MVPs, and iterating on the core problem is key.
Key Insights
The primary goal of a pre-launch startup is to launch quickly, get initial customers, talk to them for feedback, and iterate, holding the problem and customer tightly but the solution loosely.
MVPs should be ridiculously simple, buildable in weeks (not months), have extremely limited functionality, and focus on solving the highest-priority problems for a small set of initial users.
Companies like Airbnb, Twitch (as Justin TV), and Stripe (as slashdev/payments) launched with significant flaws, such as no payments or map view for Airbnb, low-resolution video and no games for Twitch, and manual payment integration for Stripe.
For industries with heavy regulation or complex hard tech, an MVP might start with a simple website explaining the product, with the 'heavy MVP' concept contrasting the 'lean MVP' that focuses on rapid software development.
Launching an MVP is less about a grand 'press launch' and more about getting any customers interacting with the product to gain real-world insights, as even major tech launches (Google, Facebook, Twitter) often go unnoticed initially.
Hacks for building an MVP quickly include time-boxing the specification, writing it down to track changes, cutting features aggressively (even important ones) to meet deadlines, and not falling in love with the MVP as it's just a starting point.
Defining the Minimum Viable Product (MVP)
An MVP is defined as the first, ridiculously simple thing you can provide to your very first target users to see if you can deliver any value at all. It is crucial to engage in some user conversations before finalizing your MVP, and even more beneficial if you are your own user. The core idea is to solve a problem that you know someone has, making the acquisition of early users straightforward—often, it's the people you already know experiencing the problem. If you're unsure who your users are, it's a sign to question the initial premise.
The essential steps for early-stage startups
The goal for any pre-launch startup is straightforward and has been a core Y Combinator ethos: Launch quickly. This means launching something that might be imperfect, and also often described as 'bad,' as soon as possible. Following this, the next critical step is to acquire initial customers and get anyone using the product. The process doesn't require a grand vision for mass adoption at this stage; simply having users interact and identify value is sufficient. Many startup journeys end before a single user has engaged with the product, so passing this hurdle is vital. The subsequent steps involve talking to any users post-launch to gather feedback and, most importantly, iterating on the product based on that feedback.
Hold the problem, not the solution, loosely
A common mistake founders make is having a fixed idea of their final product and dismissing feedback on an early, imperfect version as useless. The reality is that the grand vision in your head should remain flexible, as it might not align with customer needs. Michael Seibel advocates for holding the problem you are solving and the customer tightly, but the solution loosely. This approach allows for adaptation. For instance, if a screwdriver isn't effectively screwing in a bolt for a mechanic (the user), the problem (needing to screw something in) and the user (the mechanic) should remain constant. The failing component is the screwdriver itself, which needs to be iterated upon and fixed, rather than trying to repurpose the screwdriver into something else, like a cooking utensil.
Characteristics of a lean MVP
A lean MVP should be buildable rapidly, ideally in weeks rather than months, and can range from software applications to a simple landing page paired with a spreadsheet for initial validation. Functionality must be extremely limited, condensing only what initial users absolutely need. Founders often err by trying to address every potential user problem simultaneously. Instead, the focus should be on a small group of early users and their most pressing issues, deferring other concerns. This minimal product serves as a base for iteration, not as a definitive or special final product. It is simply a starting point.
Foundational MVPs of major tech companies
Several now-dominant tech companies began with remarkably simple and flawed MVPs. Airbnb's initial offering in 2008 lacked payment processing; users had to exchange money in person with hosts and there was no map view to locate listings. Nate, the primary coder, worked part-time. Twitch, initially Justin TV, launched as a single-channel online reality show (following Justin's life) with extremely low-resolution video and no video games; recognizing faces was difficult due to video quality. Stripe, then called slashdev/payments, had no bank deals, processed payments in a rudimentary 'startupy' way, and founders personally integrated it for early clients. These examples highlight that even billion-dollar companies started with products that were, by many standards, 'pretty shitty'.
When a 'heavy MVP' might be necessary
While the emphasis is on lean MVPs, certain industries necessitate a 'heavy MVP' due to significant regulatory hurdles or inherent complexity. Sectors like insurance, banking, or areas involving hard tech (rockets, biotech, cancer drugs), and ambitious projects (moonshots, tunneling) make rapid, lean development impossible. In such cases, the MVP might start as a simple website explaining the concept. This website acts as a reference point for interactions and can be built in days. The notion is that a 'heavy MVP' or preliminary website can sometimes paradoxically enable faster progress in complex fields than a rushed, incomplete lean attempt.
The true meaning of 'launch' for startups
Many founders have a misconception of 'launch,' often influenced by the dramatic press events of established companies. However, Seibel points out that the actual launches of major platforms like Google, Facebook, or Twitter were largely unmemorable to the public. The critical takeaway is that 'launch' for a startup should be redefined as 'getting any customers.' The goal should be to prioritize this early customer acquisition over elaborate 'press launches.' Learning from customers is exponentially more effective when they interact with a functional product rather than just discussing ideas. Product-market fit cannot be truly assessed through research alone; it requires putting a product, however basic, in front of users to immediately gauge effectiveness.
Practical hacks for rapid MVP development
To accelerate MVP development, several practical strategies can be employed. First, 'time box your spec': define a strict deadline (e.g., three weeks) and ensure the specification only includes features achievable within that timeframe, forcing the exclusion of non-essential elements. Second, 'write your spec': documenting requirements helps maintain focus and honesty about project changes. Without a written spec, it's easy to drift from the original plan without realizing it, turning a short sprint into a prolonged effort. Third, 'cut your spec': midway through development, if deadlines are at risk, aggressively prune the feature list, even removing important but non-critical items, to ensure something is delivered. Finally, 'don't fall in love with your MVP'; it's merely a foundational step. Like a first-grade paper, its significance is in what it enables next, not its intrinsic perfection.
Mentioned in This Episode
●Companies
●Organizations
●People Referenced
MVP Launch Checklist
Practical takeaways from this episode
Do This
Avoid This
Common Questions
An MVP, or Minimum Viable Product, is the absolute simplest thing you can give to your first set of target users to see if you can deliver any value to them at all. It's about delivering basic value quickly.
Topics
Mentioned in this video
Accelerator program where Michael Seibel works and which emphasizes the importance of launching MVPs quickly.
Used as a case study for an early MVP, highlighting its initial lack of payments and map view.
Presented as an example of a simple MVP, launched as Justin.tv with only one channel and low-resolution video.
The original name and concept for Twitch, which was a single online reality TV show.
Showcased as an early MVP called slashdev/payments, where founders integrated it for clients manually.
Mentioned as a company whose launch was not publicly memorable, illustrating that successful launches aren't always flashy.
Cited as an example of a company whose launch was not widely remembered, highlighting the difference between a product launch and a press launch.
Used as an example of a company whose launch was not publicly remembered, reinforcing the idea that customer acquisition is more important than a grand launch event.
More from Y Combinator
View all 562 summaries
14 minInside The Startup Reinventing The $6 Trillion Chemical Manufacturing Industry
1 minThis Is The Holy Grail Of AI
40 minIndia’s Fastest Growing AI Startup
1 minStartup School is coming to India! 🇮🇳
Found this useful? Build your knowledge library
Get AI-powered summaries of any YouTube video, podcast, or article in seconds. Save them to your personal pods and access them anytime.
Try Summify free